Experimental characterization of heterogeneous deformation due to phase transformations, twinning, and slip deformation using digital image correlation.

摘要

In traditional continuum plasticity models, deformation heterogeneities are averaged out above the length scale of a representative volume element (RVE). Below this length scale, heterogeneities are permitted but must average out over the volume of the RVE. This is the so-called homogenization technique. In recent years, the emergence of systematic experimental investigations and numerical simulations have challenged traditional continuum plasticity models, which suggest homogeneous loads produce spatially homogeneous deformations in the absence of plastic instabilities. Knowledge of strain heterogeneities is important to enhance the performance of engineering devices and components. For example, as devices and components are miniaturized, strain heterogeneities may lead to structural instability, or even product failure. Manufacturing difficulties such as those commonly found in metal forming operations which include tearing, necking, edge cracking, and surface roughening are often attributed to heterogeneous deformation. It is also speculated that strain heterogeneities may develop into fatigue damage initiation sites as well as corrosion sites.The current work is focused on characterizing the spatial strain heterogeneity due to phase transformations, twinning, and slip deformation. Using digital image correlation to obtain full-field strain measurements at multiple length scales (ranging from micrometer to millimeter), it was shown that despite uniform microstructures and boundary conditions, each deformation mechanism manifested heterogeneously. Specifically, increasing strain heterogeneity was observed concurrent with decreasing the measurement length scale. It is shown that an appropriate length scale to obtain measurements can be determined for phase transformations, and twinning deformation, but not for slip deformation. No inherent length scale could be determined for slip deformation because no discrete boundaries between slipped and un-slipped domains exist. Slip deformation appears to have a spatial correlation, and seems to obey a power-law relationship. Thus, for slip deformation, changing the length scale does not change the measurement in a one-to-one fashion whereas for phase transformations and twinning deformation it does.These multiscale measurements were used to discuss: the transformation sequence, nuances of the stress-strain curve such as strain hardening and softening, the hysteretic behavior, the strain-rate dependence, twinning strain, twin-twin intersections, twin fraction evolution, dynamic strain aging, grain interactions, mesoscopic slip bands, and a RVE size.